The present invention relates to a creation method and apparatus of network configuration for power system and more particularly to an art for deciding the connection status of a power transmission and transformation facility system of a power system of power distribution or transmission.
As a creation method of a power system configuration, it is proposed to decide a network configuration by a genetic algorithm in consideration of restrictions in the capacity of a feeder like xe2x80x9cOptimization Method of Power Distribution System Configuration by Genetic Algorithmxe2x80x9d (pp. 591 to 598) described in Journal of Institute of Electrical Engineers of Japan, Vol. C116, No. 5. According to the method, to maintain the radial network, the gene structure of the genetic algorithm is made appropriate and when a loop system or dividing system is set due to crossing-over or mutation, a countermeasure is taken individually.
The prior art does not use the operation knowledge (heruistic) on how the current network configuration is to be changed so as to improve the system, by performing a random search. As a result, when the problem scale is enlarged, it is difficult to find an operable solution (radial network) and it is more difficult to obtain an optimum solution.
If the range of neighbor status is narrowed when the space of a possible solution is divided into a plurality of spaces in a solution space, the solution space cannot be moved to another solution space unless going once via an neighbor status which is not a possible solution. On the other hand, when the range of neighbor status is widened, the search range is widened by one try. Therefore, candidates to be evaluated increase in number and the calculation time becomes longer.
A problem of the present invention is to provide a creation method and apparatus of a power network configuration which can obtain an optimum solution in a short time.
The present invention is realized on the assumption that the Tabu Search method is used as an optimization method. Recently, the Tabu Search method may be referred to the adaptive memory programming. The Tabu Search method is a method for defining a plurality of neighbor network configurations obtained by changing a part of the current network configuration (hereinafter, referred to as neighbor network configurations) and creating an optimum network configuration by repeating a transition such as moving to the neighbor network configuration in which the predetermined evaluation function becomes optimum among those neighbor network configurations and further moving sequentially to an optimum neighbor network configuration for the optimum neighbor network configuration.
However, when the local search method is applied, if the evaluation function of the system status at a certain point of iteration is worse than the evaluation function of either neighbor system status, no transition destination is found and the search ends immediately. Namely, the evaluation function cannot escape from the local minimum (minimum solution). When the evaluation function only transits to the most optimum neighbor network configuration, although it can exceed the local minimum once, it returns to the original local minimum again at the next search.
In the Tabu Search method, information (tabu tenure) changed with search is stored in a finite length taboo list, and movement toward the taboo stored in the taboo list is inhibited, and the evaluation function will not return to the local minimum. Namely, since movement toward the tabooed direction is inhibited for a given period, a wide solution space can be searched most widely and the local search can be speeded up. Since the Tabu Search method defines the neighbor network configuration appropriately using the heuristic method, it is possible to omit a useless search and obtain a better solution at a higher speed.
When although there is a better solution than the optimum solution obtained up to a certain point of iteration in the neighbor network configuration, there is a taboo for inhibiting movement to the solution in the taboo list, a mechanism (aspiration level) that movement is permitted though it is tabooed and the optimum solution in the neighborhood is searched can be introduced.
From the aforementioned viewpoint, the present invention is characterized in that one of a plurality of switches constituting a power system is set as a perceived switch, and n optional switches Bk (k=1 to n) in the on state among the other switches are changed to the off state, and n loop networks including the perceived switch are created, and each switch Ck (k=1 to n) constituting the loop networks is changed from the off state to the on state so as to create a plurality of network configuration candidates, and a power network configuration is decided on the basis of these plurality of network configuration candidates.
The power network configuration is decided by evaluating a plurality of network configuration candidates on the basis of the predetermined evaluation standard. The evaluation standard can use an evaluation function having at least one of the power transmission loss, the difference between the power flow restriction value and the power flow value, the ratio of the power flow restriction value to the power flow value, the number of facilities of power flow restriction violation, the difference between the voltage lower limit and the voltage, the ratio of the voltage lower limit to the voltage, and the number of facilities of voltage restriction violation as an item.
The present invention is based on that for example, each facility constituting a power system is expressed by a node and branch on the basis of the connection relationship, and an optional facility is perceived, and a switch for changing the power operation conditions (power transmission loss, restrictions of tidal current, restrictions of voltage, etc.) of the branch or node which is the perceived facility is aimed at, and by defining the network configuration obtained by turning the switch on or off under the condition of maintaining the radial network configuration as an neighbor network configuration, useless searches are reduced and the searching process is speeded up. Furthermore, by making the range of neighbor network configuration variable, an optimum power network configuration can be created even in a discontinuous solution space.
In other words, by turning the switch on so as to create a loop system including the perceived branch and turning another switch included in the created loop on so as to open the loop, the status amount (voltage, power flow, power transmission loss, etc.) of the perceived facility can be changed. At the same time, by using the heuristic method for always maintaining a radial power system, neighbor network configurations for changing the status amount of power operation conditions, that is, candidates of network configuration to be created can be created. By repeating transition to the optimum network configuration among the neighbor network configurations by the Tabu Search method, useless searches can be reduced much more. Furthermore, since the search area which is a possible solution can be searched widely, an optimum power network configuration can be created.